CN106407235B - A kind of semantic dictionary construction method based on comment data - Google Patents
A kind of semantic dictionary construction method based on comment data Download PDFInfo
- Publication number
- CN106407235B CN106407235B CN201510469211.4A CN201510469211A CN106407235B CN 106407235 B CN106407235 B CN 106407235B CN 201510469211 A CN201510469211 A CN 201510469211A CN 106407235 B CN106407235 B CN 106407235B
- Authority
- CN
- China
- Prior art keywords
- word
- semantic
- template
- dictionary
- pat
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000010276 construction Methods 0.000 title claims abstract description 11
- 239000000284 extract Substances 0.000 claims abstract description 9
- 238000000034 method Methods 0.000 claims description 38
- 230000008451 emotion Effects 0.000 claims description 21
- 238000011156 evaluation Methods 0.000 claims description 16
- 238000000605 extraction Methods 0.000 claims description 11
- 230000011218 segmentation Effects 0.000 claims description 9
- 238000002372 labelling Methods 0.000 claims description 8
- 238000004364 calculation method Methods 0.000 claims description 7
- 238000003780 insertion Methods 0.000 claims description 4
- 230000037431 insertion Effects 0.000 claims description 4
- 230000004927 fusion Effects 0.000 claims description 3
- 238000012545 processing Methods 0.000 claims description 3
- 238000004458 analytical method Methods 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 3
- 238000011430 maximum method Methods 0.000 description 3
- 238000003058 natural language processing Methods 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 101100340610 Mus musculus Igdcc3 gene Proteins 0.000 description 1
- 235000021152 breakfast Nutrition 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 239000000470 constituent Substances 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 230000002996 emotional effect Effects 0.000 description 1
- 238000004880 explosion Methods 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 239000004615 ingredient Substances 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000000452 restraining effect Effects 0.000 description 1
- 238000012552 review Methods 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000013106 supervised machine learning method Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
- 239000002023 wood Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/36—Creation of semantic tools, e.g. ontology or thesauri
- G06F16/374—Thesaurus
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Machine Translation (AREA)
Abstract
The present invention relates to a kind of semantic dictionary construction method based on comment data, step includes: 1) to construct seed semantic dictionary by comment data on a small quantity；2) comment data are segmented；3) semantic category of data is commented on by word judgment and be replaced with semantic category label；4) template is generated according to the concrete term that the title of each semantic category and each semantic category include；5) in the comment data after template to be applied to semantic category tag replacement, to extract the semantic word of each semantic category；6) it is given a mark according to the importance of template, generalization and accuracy to each template；7) the part template for choosing highest scoring calculates the score for the semantic word that each template extracts, and then the part of semantic word for choosing highest scoring expands semantic dictionary；8) step 3)~7) iteration progress, final semantic dictionary and template library are obtained after termination.The present invention can obtain fairly large semantic dictionary within a short period of time, and can extract multiple semantic categories simultaneously.
Description
Technical field
The invention belongs to information technologies, data mining technology field, and in particular to a kind of semantic word based on comment data
Allusion quotation construction method.
Background technique
With the fast development of e-commerce, the comment on internet is used from the people visual field is progressed into slowly influence
The selection at family, then just deepening step by step to the influence to brand.By taking hotel industry as an example, hotel, which wishes to obtain by technological means, to be used
The comment at family is fed back, and for instructing the Brand management and operation management in hotel, promotes brand image and service quality.User wishes
The comment for checking other users, the advantages of specifying hotel and disadvantage, in this, as the important references of reservation.Tripadvisor is ground
Study carefully display, the user more than 85% pays much attention to the public praise quality in hotel, and nearly 90% user checks before making reservation decision
User reviews.
More and more users are happy to share oneself viewpoint or experience on the internet, and this kind of comment data explosion formula increases
Long, only method manually is difficult to cope with the collection and processing of online magnanimity comment.Therefore, there is an urgent need to computer help users
Quick obtaining comes into being with these comment information, sentiment analysis (Sentiment Analysis) technology is arranged.Sentiment analysis
It is not only the research hotspot of field of information processing, also results in extensive concern in industrial circle.
The emotion for analyzing comment first has to identify the valuable emotion information element in comment, this includes: 1) to comment
Valence object, such as " hotel ", " price "；2) evaluative component, such as " very good ", " can be said to be clean ".Wherein, evaluative component includes
Emotion word (such as " good ", " clean "), degree adverb (such as " very "), common adverbial word (such as " mostly ") and negative word
(such as " no "), evaluative component not only expresses emotion, is also reinforced by its ornamental equivalent, weakens or set anti-emotional expression sentence
Feeling polarities, so that it is more abundant to obtain emotional expression.
Importance of the emotion word in sentiment analysis is self-evident.However in many cases, individual emotion word
Polarity be it is ambiguous, such as "high" of " price in dining room is very high " describes to indicate derogatory sense when " dining room price ", and " restaurant employee's work
It is very high to make efficiency " "high" description " working efficiency " when indicate commendation.Therefore, emotion is only considered in the sentiment analysis of text
Word is far from being enough, it is also necessary to consider the collocation of evaluation object and emotion word, such as<price, high>,<working efficiency, high>in this way
Binary collocation.
Above-mentioned all kinds of semantic dictionaries, either emotion word dictionary, degree adverb dictionary etc. or dictionary of collocations etc., for
Text emotion analysis plays the role of very important.The current pure dictionary resources for artificially collecting arrangement, scale is inadequate, efficiency
Also very low.A kind of better method is statistical method or machine learning method based on corpus, although this method can band
Carry out some noises, but at this moment intervene again manually, cost is relatively low.Bootstrapping is a kind of semi-supervised engineering
Learning method is widely applied in information extraction, construction of knowledge base field, can also be used for reference and is applied in semantic dictionary building.
Summary of the invention
The present invention in view of the above-mentioned problems, provide it is a kind of based on comment data semantic dictionary construction method, height can be generated
The semantic dictionary and template library of quality.
The technical solution adopted by the invention is as follows:
A kind of semantic dictionary construction method based on comment data, includes the following steps:
1) comment data are obtained, the semantic word of each semantic category is obtained by commenting on data on a small quantity, construct seed semanteme word
Allusion quotation；
2) word segmentation processing is carried out to the sentence of comment data；
3) it to the comment data after participle, is replaced by its semantic category of word judgment and with semantic category label；
4) make pauses in reading unpunctuated ancient writings to the comment data after tag replacement, the tool for including according to the title of each semantic category and each semantic category
Pronouns, general term for nouns, numerals and measure words language generates template；
5) in the comment data after template to be applied to semantic category tag replacement, to extract the semantic word of each semantic category；
6) it according to the importance of template, generalization and accuracy, gives a mark to each template；
7) the part template for choosing highest scoring calculates the semantic word that each template extracts according to the template of selection and its marking
Score, and then choose highest scoring part of semantic word semantic dictionary is expanded；
8) step 3) to step 7) iteration carries out, and iteration ends when select semanteme word is incorrect obtain most
Whole semantic dictionary, and template library is constituted by each template.
Further, step 1) obtains online comment data from comment website by focused crawler, and by manually checking
A small amount of comment, arranges the semantic word of each semantic category, forms seed dictionary.
Further, step 2) is segmented using the maximum match segmentation based on dictionary first, is then directed to and is divided
The ambiguous part of word obtains correct word segmentation result using the segmenting method of sequence labelling；The segmenting method of the sequence labelling
The cutting problems of word are converted to the classification problem of word, each radical assigns different positions according to its different location in word
Category label determines the slit mode of sentence based on such flag sequence.
Further, the step 3) semantic category include evaluation object word, it is evaluation attributes word, emotion word, degree adverb, general
Logical adverbial word, negative word, insertion word.
Further, step 4) according to ".","！", "? " 3 punctuation marks are made pauses in reading unpunctuated ancient writings, and the minimum for limiting template is long
Degree is 3 words, and maximum length is 7 words.
Further, when step 5) extracts the semantic word of each semantic category, when some corresponding template of comment segment and step
4) when only one word of difference of gained template, using the word as the example word of corresponding semantic category.
Further, the part template of the step 7) highest scoring is preceding 5~10% template of highest scoring, described
The part of semantic word of highest scoring is preceding 5~10% semantic word of highest scoring.
Further, after step 8), by the polarity and emotion word that are manually determined emotion word in semantic dictionary
With the collocation polarity of evaluation object word, evaluation attributes word；In artificial determination process, by the corresponding comment segment work of its affiliated template
For the foundation of judgement.
Compared with the pure mode artificially collected, the present invention use based on comment corpus method it is high-efficient, can compared with
It is arranged in short time and obtains fairly large semantic dictionary；Compared with traditional Bootstrapping method, mould proposed by the present invention
Version marking can effectively measure the situation of template nesting；Semantic dictionary construction method phase with tradition based on Bootstrapping
Than the present invention can extract multiple semantic categories simultaneously.
Detailed description of the invention
Fig. 1 is the step flow chart of the semantic dictionary construction method of the invention based on comment data.
Specific embodiment
In order to make the foregoing objectives, features and advantages of the present invention clearer and more comprehensible, below by specific embodiment and
Attached drawing, the present invention will be further described.
Building for semantic dictionary, the present invention use the method based on bootstrapping (Bootstrapping).Bootstrapping, i.e., certainly
Extension or bootstrapping are a kind of semi-supervised machine learning methods, can be used for extracting semantic dictionary and template simultaneously.This method
Thought based on the observation that extraction template can be used for extracting new example, these examples can be used for taking out again in turn
Take new template.The advantage of this method is not needing the training corpus of mark, it is thus only necessary to a small number of seeds.Pass through people first
Work intervenes the seed word initialized, obtains template using seed word, and then obtain seed word by template, so
Iteration carries out.In each round iteration, new labeled data will be all generated, optimal word can be added to accordingly to semantic dictionary
In, optimal template can also be added in template library, model be relearned with these new labeled data, to can produce again
New data, loop back and forth like this, and terminate until finally restraining, to obtain more seed words and template.Here it is most
Basic Bootstrapping algorithm (or process).
The semantic category of semantic dictionary includes evaluation object word (Obj), evaluation attributes word (Attr), emotion word (Sent), journey
Adverbial word (Dgr), common adverbial word (Adv), negative word (Neg), insertion word (Inter) etc. are spent, each semantic category includes several words
Language, template are exactly the sequence being made of semantic class name or concrete term.
The step of present invention improves on the basis of existing Bootstrapping method, and Fig. 1 is the method for the present invention stream
Cheng Tu, here are specific implementation steps:
Step 1: data preparation.The online comment data of website acquisition are commented on from the mainstreams such as journey are taken by focused crawler.
Step 2: seed dictionary creation.It manually checks a small amount of (such as 500) comment, arranges the semantic word of each semantic category,
Seed dictionary is formed, which is denoted as SemLex.
Step 3: comment participle.Chinese word segmentation is the basic steps of Chinese natural language processing, and present invention participle uses word
The method of allusion quotation participle and statistics participle fusion.The maximum match segmentation based on dictionary is used first, has ambiguity for participle
Part use the segmenting method of sequence labelling again.
Maximum match segmentation based on dictionary gives dictionary, for chinese character sequence to be segmented, successively finds matching
Longest dictionary word, no matcher is then used as monosyllabic word to handle, until the chinese character sequence is disposed.It is swept according to chinese character sequence
Retouch the difference in direction, this method can be divided into again: Forward Maximum Method (matching from left to right) and reverse maximum matching are (from dextrad
Left matching).For example, for sequence " when the atom binding constituents period of the day from 11 p.m. to 1 a.m ", Forward Maximum Method result be " when | atom | in conjunction with | at |
Molecule | when ", and reverse maximum matching result is " when | atom | in conjunction with | ingredient | the period of the day from 11 p.m. to 1 a.m ".
Obviously, Forward Maximum Method and reverse maximum matching cannot all handle cutting ambiguity problem well.It is positive maximum
Matching and reverse maximum matching also may be combined to form two-way maximum matching, forward direction and inversely match inconsistent when bi-directional matching
Place, the often place of potential ambiguity.There is ambiguity to generally require to confirm word segmentation result according to specific context.There is the sequence of supervision
Column mask method can adequately excavate the feature-rich of context, therefore present invention introduces sequence labellings in ambiguous situation
Method disambiguation.The cutting problems of word are converted to the classification problem of word by this method, and each radical is according to its difference in word
Position assigns different position classification labels, for example, in prefix, word, suffix and monosyllabic word.Based on such flag sequence, very
It is easy to determine the slit mode of sentence.Wherein, B (Begin), M (Middle), E (End), S (Single) respectively indicate prefix,
In word, suffix, monosyllabic word.There is the flag sequence of word, the word sequence for meeting regular expression " S " or " B (M) * E " indicates one
Word, to be readily accomplished sentence cutting.In order to realize that sequence labelling task, the present invention use conditional random field models
(Conditional Random Fields, CRF), which is used widely in natural language processing, and achieves very
Ten-strike.Specific features include: previous word, current word, the latter word, previous word and current word, current word and the latter
Word, and the binary feature based on these unitary features.For conditional random field models using these features extracted, what is predicted is every
The category label of a word.
The dictionary of maximum matching process and have supervision conditional random field models training study corpus both be from this hair
Bright 100,000 manually marked hotel comment.
Step 4: semantic category tag replacement.It is replaced to the comment after participle by its semantic category of word judgment and with semantic category label
It changes, as " dining room | | price | very | it is high ", replace with " Obj | | Attr | Dgr | Sent ", for commenting on starting and ending position
Add " Start " and " End " label respectively, in comment in addition to ".","！", "? " except punctuation mark also use " Punc " mark
Label replacement.
Step 5: template generates.The step makes pauses in reading unpunctuated ancient writings to the comment data after tag replacement, according to the name of each semantic category
The concrete term that title and each semantic category include generates template.In the present embodiment, according to ".","！", "? " 3 punctuation mark punctuates,
3 words of template minimum length, 7 words of maximum length are limited, the comment after scanning tag replacement generates candidate template.
Step 6: semantic word extracts.In comment after candidate template to be applied to semantic category tag replacement.When some comment
When only one word of difference of the corresponding template of segment and candidate template, using the word as the example word of corresponding semantic category.For example,
For comment segment " price | very | high ", wherein " price " belongs to evaluation attributes word, "high" belongs to emotion word, and is not belonging to " very much "
Any semantic category, at this moment its corresponding template is " Attr | very | Sent ".This with candidate template " Attr | Dgr | Sent " only in
Between a word difference, then will extract " very much " the example word as degree adverb.
Step 7: template marking.The present invention gives a mark in terms of two, on the one hand measures the importance of template by the frequency and pushes away
On the other hand wide property measures the accuracy of template by the hit rate in semantic dictionary.
Template patiImportance and generalization marking S (pati) calculation formula it is as follows:
Wherein, | pati| it is template patiLength, with word number calculating, f (pati) indicate template patiThe frequency, C
(pati) indicate nesting patiTemplate set, as comment segment " dining room | | price | very | it is high " corresponding template " Obj | |
Attr | Dgr | the corresponding template of the nested comment segment of Sent " " price | very | high " " Attr | Dgr | Sent ".
patiAccuracy marking P (pati) calculation formula it is as follows:
Wherein, T (pati) indicate template patiThe semantic set of words of extraction, f (t) indicate the frequency of semantic word t, SemLex
The seed semantic dictionary constructed for step 1.
We use Sigmoid functionBy S (pati) normalize to (0,1), and then merge two aspects
Marking obtain F (pati), calculation formula is as follows:
Wherein α is importance and generalization marking S (pati) weight, value range be [0,1].The present invention more focuses on mould
The accuracy of version, therefore by α=0.4, it can also be adjusted according to concrete application.
Step 8: template is selected.According to F (pati) choose highest scoring preceding 5~10% template.
Step 9: semantic word marking.According to the template pat selectedkAnd its marking, calculate the semantic word of template extraction
Score, calculation formula are as follows:
Step 10: semantic dictionary expands.Preceding 5~10% word for choosing highest scoring is added to semantic dictionary SemLex
In.
Step 4 is carried out to step 10 iteration.Stopping criterion for iteration.It is select semanteme word it is obviously incorrect when terminate.
Step 11: polarity determines.Polarity and emotion word for emotion word and evaluation object word, evaluation attributes word
Collocation polarity, by being accomplished manually.In artificial determination process, using the corresponding segment of commenting on of its affiliated template as the foundation determined.
The result shows that the present invention achieves good performance in accuracy rate and recall rate, the semanteme of high quality is generated
Dictionary and template library.
It is in the comment of 10,000,000 hotels the experimental results showed that, semantic dictionary construction method proposed by the present invention is effective
's.The evaluation object word of extraction has 4835, such as " breakfast ", " network "；The evaluation attributes word of extraction has 175, such as " valence
Lattice ", " attitude " etc.；The emotion word of extraction has 2393, such as " comfortable ", " praising "；The degree adverb of extraction has 92, and such as " ten
Point ", " excessive " etc.；The common adverbial word extracted has 214, such as " very ", " excessive "；The negative word of extraction has 28, such as " wood
Have ", " will not " etc.；The insertion word of extraction has 143, such as " feeling ", " generally speaking ".
The above embodiments are merely illustrative of the technical solutions of the present invention rather than is limited, the ordinary skill of this field
Personnel can be with modification or equivalent replacement of the technical solution of the present invention are made, without departing from the spirit and scope of the present invention, this
The protection scope of invention should be subject to described in claims.
Claims (10)
1. a kind of semantic dictionary construction method based on comment data, which comprises the steps of:
1) comment data are obtained, the word of each semantic category is obtained by commenting on data on a small quantity, construct seed semantic dictionary；
2) word segmentation processing is carried out to the sentence of comment data；
3) it to the comment data after participle, is replaced by its semantic category of word judgment and with semantic category label；
4) make pauses in reading unpunctuated ancient writings to the comment data after tag replacement, the specific word for including according to the title of each semantic category and each semantic category
Language generates template；
5) in the comment data after template to be applied to semantic category tag replacement, to extract the semantic word of each semantic category；
6) it according to the importance of template, generalization and accuracy, gives a mark to each template；
7) the part template for choosing highest scoring calculates obtaining for the semantic word that each template extracts according to the template of selection and its marking
Point, and then the part of semantic word for choosing highest scoring expands semantic dictionary；
8) step 3) to step 7) iteration carries out, and iteration ends when select semanteme word is incorrect obtain final
Semantic dictionary, and template library is constituted by each template,
Wherein, the method that step 6) gives a mark to each template is:
A) to template importance and generalization marking S (pati) calculation formula it is as follows:
Wherein, | pati| it is template patiLength, with word number calculating, f (pati) indicate template pati the frequency, C (pati) table
Show nested patiTemplate set；
B) to template accuracy marking P (pati) calculation formula it is as follows:
Wherein, T (pati) indicate template patiThe semantic set of words of extraction, f (t) indicate the frequency of semantic word t, and SemLex is kind
Sub- semantic dictionary；
C) fusion steps a), both sides marking b) obtained by the way of weighting.
2. the method as described in claim 1, it is characterised in that: step 1) obtains online point from comment website by focused crawler
Data are commented, and by manually checking a small amount of comment, arranges the word of each semantic category, forms seed dictionary.
3. the method as described in claim 1, it is characterised in that: step 2) is first using the maximum matching participle side based on dictionary
Method is segmented, and then obtains correct word segmentation result using the segmenting method of sequence labelling for the ambiguous part of participle；
The cutting problems of word are converted to the classification problem of word by the segmenting method of the sequence labelling, each radical according to its in word not
Same position is assigned different position classification labels, the slit mode of sentence is determined based on such flag sequence.
4. method as claimed in claim 3, it is characterised in that: the different position classification label, including in prefix, word,
Suffix and monosyllabic word, and sequence labelling task is realized using conditional random field models.
5. the method as described in claim 1, it is characterised in that: the step 3) semantic category includes evaluation object word, evaluation category
Property word, emotion word, degree adverb, common adverbial word, negative word, insertion word.
6. the method as described in claim 1, it is characterised in that: step 4) basis ".","！", "? " 3 punctuation marks break
Sentence, and the minimum length of template is limited as 3 words, maximum length is 7 words.
7. the method as described in claim 1, it is characterised in that: when step 5) extracts the semantic word of each semantic category, when some point
When commenting only one word of difference of template obtained by the corresponding template of segment and step 4), using the word as the example of corresponding semantic category
Word.
8. the method as described in claim 1, which is characterized in that the fusion steps a) by the way of weighting, b) obtain
Both sides marking, comprising:
Using Sigmoid functionBy S (pati) normalize to (0,1), and then merge both sides and give a mark
To F (pati), calculation formula is as follows:
Wherein α is importance and generalization marking S (pati) weight, value range be [0,1].
9. the method as described in claim 1, it is characterised in that: the part template of the step 7) highest scoring is highest scoring
Preceding 5~10% template, the part of semantic word of the highest scoring is preceding 5~10% semantic word of highest scoring.
10. the method as described in claim 1, it is characterised in that: after step 8), by being manually determined in semantic dictionary
Polarity and emotion word and evaluation object word, the collocation polarity of evaluation attributes word of emotion word；In artificial determination process, by it
The corresponding segment of commenting on of affiliated template is as the foundation determined.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510469211.4A CN106407235B (en) | 2015-08-03 | 2015-08-03 | A kind of semantic dictionary construction method based on comment data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510469211.4A CN106407235B (en) | 2015-08-03 | 2015-08-03 | A kind of semantic dictionary construction method based on comment data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106407235A CN106407235A (en) | 2017-02-15 |
| CN106407235B true CN106407235B (en) | 2019-06-11 |
Family
ID=58008194
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510469211.4A Active CN106407235B (en) | 2015-08-03 | 2015-08-03 | A kind of semantic dictionary construction method based on comment data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106407235B (en) |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107122471B (en) * | 2017-05-02 | 2020-07-10 | 北京众荟信息技术股份有限公司 | Hotel characteristic comment extraction method |
| EP3642733A4 (en) * | 2017-07-31 | 2020-07-22 | Beijing Didi Infinity Technology and Development Co., Ltd. | System and method for segmenting a sentence |
| CN108304373B (en) * | 2017-10-13 | 2021-07-09 | 腾讯科技（深圳）有限公司 | Semantic dictionary construction method and device, storage medium and electronic device |
| CN109033082B (en) * | 2018-07-19 | 2022-06-10 | 深圳创维数字技术有限公司 | Learning training method and device of semantic model and computer readable storage medium |
| CN109446310B (en) * | 2018-10-30 | 2020-11-03 | 腾讯科技（武汉）有限公司 | Question template quality evaluation method and device and storage medium |
| CN110674260B (en) * | 2019-09-27 | 2022-05-24 | 北京百度网讯科技有限公司 | Training method and device of semantic similarity model, electronic equipment and storage medium |
| CN111178045A (en) * | 2019-10-14 | 2020-05-19 | 深圳软通动力信息技术有限公司 | Automatic construction method of non-supervised Chinese semantic concept dictionary based on field, electronic equipment and storage medium |
| CN111062216B (en) * | 2019-12-18 | 2021-11-23 | 腾讯科技（深圳）有限公司 | Named entity identification method, device, terminal and readable medium |
| CN111325018B (en) * | 2020-01-21 | 2023-08-11 | 上海恒企教育培训有限公司 | Domain dictionary construction method based on web retrieval and new word discovery |
| CN112711941B (en) * | 2021-01-08 | 2022-12-27 | 浪潮云信息技术股份公司 | Emotional score analysis processing method based on emotional dictionary entity |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9213704B2 (en) * | 2010-09-20 | 2015-12-15 | Microsoft Technology Licensing, Llc | Dictionary service |
| US9552374B2 (en) * | 2013-08-19 | 2017-01-24 | Kodak Alaris, Inc. | Imaging workflow using facial and non-facial features |
| CN103886053A (en) * | 2014-03-13 | 2014-06-25 | 电子科技大学 | Knowledge base construction method based on short text comments |
-
2015
- 2015-08-03 CN CN201510469211.4A patent/CN106407235B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN106407235A (en) | 2017-02-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106407235B (en) | A kind of semantic dictionary construction method based on comment data | |
| CN106407236B (en) | A kind of emotion tendency detection method towards comment data | |
| CN106649818B (en) | Application search intention identification method and device, application search method and server | |
| WO2021114745A1 (en) | Named entity recognition method employing affix perception for use in social media | |
| CN104809176B (en) | Tibetan language entity relation extraction method | |
| CN107797991B (en) | Dependency syntax tree-based knowledge graph expansion method and system | |
| CN104268160B (en) | A kind of OpinionTargetsExtraction Identification method based on domain lexicon and semantic role | |
| CN104933039B (en) | Resourceoriented lacks the entity link system of language | |
| CN105205699A (en) | User label and hotel label matching method and device based on hotel comments | |
| CN103049435B (en) | Text fine granularity sentiment analysis method and device | |
| CN108628833B (en) | Method and device for determining summary of original content and method and device for recommending original content | |
| CN107180025B (en) | Method and device for identifying new words | |
| CN100595760C (en) | Method for gaining oral vocabulary entry, device and input method system thereof | |
| CN108681537A (en) | Chinese entity linking method based on neural network and word vector | |
| CN103077164A (en) | Text analysis method and text analyzer | |
| CN108062304A (en) | A kind of sentiment analysis method of the comment on commodity data based on machine learning | |
| Shimada et al. | Analyzing tourism information on twitter for a local city | |
| CN104008091A (en) | Sentiment value based web text sentiment analysis method | |
| CN104298665A (en) | Identification method and device of evaluation objects of Chinese texts | |
| CN102968408A (en) | Method for identifying substance features of customer reviews | |
| CN101556596B (en) | Input method system and intelligent word making method | |
| CN104778256A (en) | Rapid incremental clustering method for domain question-answering system consultations | |
| CN107092605A (en) | A kind of entity link method and device | |
| CN109165273A (en) | General Chinese address matching method facing big data environment | |
| CN110134934A (en) | Text emotion analysis method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information |
Address after: 100088 Madian East Road, Haidian District, No. 17,, golden floor, International Building, 18 Applicant after: BEIJING JOINT WISDOM INFORMATION TECHNOLOGY CO.,LTD. Address before: 100088 Beijing, Madian, East Haidian District Road, room 17, room 15, level 1818 Applicant before: BEIJING ZHONGHUI INFORMATION TECHNOLOGY Co.,Ltd. |
|
| CB03 | Change of inventor or designer information | ||
| CB03 | Change of inventor or designer information |
Inventor after: Lin Xiaojun Inventor after: Zhang Meng Inventor after: Bao Xiao Inventor after: Jiao Yu Inventor before: Lin Xiaojun Inventor before: Zhang Meng Inventor before: Bao Xiao |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20231108 Address after: 18th Floor, Jin'ao International Building, No. 17 Madian East Road, Haidian District, Beijing, 100080 Patentee after: Beijing Yishang Huiping Network Technology Co.,Ltd. Address before: 100088 18 / F, jin'ao international building, 17 Madian East Road, Haidian District, Beijing Patentee before: BEIJING JOINT WISDOM INFORMATION TECHNOLOGY CO.,LTD. |